Over the last few years, I've had the opportunity to engage with more than a hundred vendors of big data solutions. While each has its own distinct view of the market and its own particular offerings, most can be readily understood after a quick conversation. For example, I spoke with a startup company today, Cambridge Intelligence, that has a Java developer toolkit for visualizing graph analytics in browser-based applications—pretty straightforward, even if they don't have a big marketing department to communicate this broadly. Conversely, sometimes the bigger players are the hardest to pin down precisely, if only because they have so many more products and organizational boundaries to reconcile.
One outlier in the other direction is HP. Despite the cute HAVEn acronym, it's hard to find a unifying theme in HP's approach to big data. The HAVEn concept itself seems more like a list of stuff (Hadoop, Autonomy, Vertica, Enterprise Security, and n-applications) than a cohesive platform. There has been much talk about the integration of Vertica's SQL data warehouse capabilities with Autonomy's IDOL search and analytics engine and with their Distributed R flavor of the popular analytics language, but this still seems more marketing than technology. Oh, and these components have some connectors to Hadoop/HDFS. Plus the security thing—what was the security thing again?
HP is hosting its big data conference in Boston this week, so it will be interesting to read the announcements and read between the lines to see if their strategy is more (or less) than the sum of its parts. I, for one, would welcome more clarity about how potential customers would actually work with HP to consume these solutions in a coherent fashion to build real big data solutions.